The associations among platelet count, mean platelet volume, and erectile dysfunction: an observational and Mendelian randomization study

Abstract

Background

While previous studies have explored the associations and causalities among platelet count (PC), mean platelet volume (MPV), and erectile dysfunction (ED), further investigations are needed to clarify these relationships using advanced methodologies and analyzing specific populations.

Aim

To investigate the associations and causalities among PC, MPV, and ED using observational study and Mendelian randomization (MR) analysis.

Methods

A total of 114 patients with ED and 158 healthy control participants underwent a fasting blood draw to test for PC and MPV along with a comprehensive laboratory examination. The International Erectile Function Index was used to diagnose ED. Genetic variants of ED were obtained from individuals of European ancestry including 6175 ED cases and 217 630 controls. PC and MPV values were obtained from the UK Biobank and Investigating the effect of varying the whole blood inter-donation interval (INTERVAL) studies, encompassing a cohort of 173 480 individuals of European descent. Inverse-variant weighted (IVW), weighted median (WM), and MR-Egger methods were employed in MR analysis to explore the causal effects between variables to assess the impact of PC and MPV on ED. Various sensitivity analyses were employed to ensure the reliability of the results.

Outcomes

Both observational study results and MR results revealed that elevated PC levels were associated with a heightened risk of ED, whereas reductions in MPV were linked to a decreased risk.

Results

Logistic regression analysis indicated that an increased PC was associated with a greater risk of ED, with an odds ratio (OR) of 1.14 (95% CI: 1.08, 1.22; P = .005), whereas decreased MPV was linked to an increased risk for ED, with an OR of 0.65 (95% CI: 0.48, 0.88; P = .003). Our MR analysis also revealed that genetically predicted PC was associated with a 1.09-fold increased risk of ED (95% CI: 1.01, 1.18; P = .016). Conversely, genetically predicted MPV was linked to a 0.93-fold increased risk of ED (95% CI: 0.88, 0.99; P = .014). The absence of heterogeneity (P > .05) and pleiotropy (P > .05) was confirmed through Cochran's Q tests and MR-Egger regression. Exclusion of individual single-nucleotide polymorphisms (SNPs) did not alter the robustness of the results.

Clinical implications

In clinical work, it is an important guide for the prevention, diagnosis, and treatment of ED.

Strengths and limitations

Our study employed a combination of observational studies and MR studies to strengthen our evidence. The observational study's sample size was relatively small, and MR was limited to individuals of European ancestry.

Conclusion

A high PC and a low MPV are associated with an increased risk of ED, highlighting the importance of addressing platelet parameters in ED management.

Introduction

The National Institutes of Health defines erectile dysfunction (ED) as the inability to achieve and sustain an erection sufficient for satisfactory sexual intercourse.¹ Erectile dysfunction is typically categorized into 3 main classifications: organic, psychogenic, and mixed.² ED is the most common form of sexual dysfunction, often negatively impacting patients' relationship and quality of life, as well as their partners.³ Epidemiologic data indicate that the prevalence and incidence of ED are high and increasing worldwide.^2,4 The number of patients with ED is predicted to reach 322 million worldwide in 2025, which places a heavy burden on health care, finance, and society.⁵

Vascular endothelial impairment is a common cause of organic ED.⁶ Platelets, in their physiologic state, have the ability to decrease blood outflow from the penile veins by enhancing the hypercoagulable state of penile blood flow in the initial phase of penile erection, thereby facilitating increased penile blood flow, which is beneficial for penile erection.⁷ In the pathological state, the impairment of vascular endothelial cells triggers platelet activation, leading to altered function. The release of inflammatory transmitters and cytokines subsequently promotes and accelerates the progression of atherosclerosis, contributing to the development of ED.⁸

The relationship between platelet indices and ED has become a topic of interest in recent research, establishing a significant correlation between the mean platelet volume (MPV) and ED. A meta-analysis of 14 studies revealed that individuals with ED, particularly those with vasculogenic ED, have higher MPV levels than healthy controls do.⁹ Another study confirmed these findings, showing a standardized mean difference in MPV between ED patients and controls. However, no significant differences were observed in the platelet distribution width (PDW) or platelet count (PC) between the ED group and the control group.¹⁰ However, the associations between PC, MPV, and ED require further elucidation due to the relatively small sample sizes of the studies included in the meta-analysis and the existing controversies in prior observational studies.

To address the confounders and reverse causality between PC, MPV, and ED, we adopted a novel framework integrating an observational study and a Mendelian randomization (MR) study. Mendelian randomization is an epidemiological method that uses genetic variants to determine the causality between exposure factors and outcome risks.^11,12 The randomly distributed single-nucleotide polymorphisms (SNPs) are used as instrumental variables (IVs) to replace the exposures (ie, PC, MPV) and outcomes (ie, ED).

In this study, we triangulated observational and MR studies to investigate the associations and causations of ED with PC and MPV.

Materials and methods

Study design

In this study, we used observational and MR studies to assess the association and causation of ED with PC and MPV (Figure 1). Using data from the Third Xiangya Hospital of Central South University, we conducted cross-sectional analysis to examine the associations of PC and MPV of ED. Furthermore, we conducted two sample MR studies, with the associations between the genetic variants and exposure and between the variants and outcome estimated from 2 sets of individuals, to confirm the causal associations of ED with PC and MPV.

Participants

All patients provided informed consent in accordance with the institutional review board guidelines and the Declaration of Helsinki. All protocols were approved by the Third Xiangya Hospital institutional review board of Central South University (No. 2019S252). From January to May 2019, the patients underwent a physical examination at the Third Xiangya Hospital's Health Management Center. The participants were outpatients and provided informed consent, with the option to pause or withdraw from interviews at any point. For the ED group, the inclusion criteria were as follows: (1) 18-80 years of age; (2) the International Erectile Function Index (IIEF-5) score was suggestive of ED (IIEF-5 score <21); (3) no other ED-related illnesses; and (4) willingness to take part in the research. The non-ED group inclusion criteria were as follows: (1) No history of ED; (2) the IIEF5 score indicates non-ED (IIEF-5 score > 21); (3) no other diseases (cardiovascular disease, diabetes, or serious diseases of other systems, such as severe cerebrovascular diseases, liver, and kidney insufficiency, blood diseases, mental diseases, malignant tumours, and so on); and (4) willingness to participate in study. This study included 272 participants, including 114 patients with ED and 158 patients without ED.

Measures

All participants underwent physical examination, which included measurements of weight, height, blood pressure, pulse, breathing, body mass index (BMI), educational background, smoking and drinking habits, and so on. All hematologic parameters were obtained from the hospital laboratory. We assessed sexual function via andrology-related scales such as the IIEF-5 and the Premature Ejaculation Diagnostic Tool (PEDT). The IIEF-5 questionnaire is used to diagnose erectile dysfunction, whereas the PEDT is used for premature ejaculation. We utilized the International Prostate Symptom Score (IPSS) to assess prostatitis. Patients with an IIEF-5 score ranging from 21 to 25 points were included in the non-ED group, whereas patients with an IIEF-5 score ranging from 0 to 20 points were included in the ED group. By using the PEDT, 3 types of premature ejaculation (PE) were diagnosed: PE (≥11), suspected PE (9-10), and non-PE (≤8). A Generalized Anxiety Disorder-7 (GAD-7), which assesses anxiety symptoms, and a Patient Health Questionnaire-9 (PHQ-9), which assesses depression symptoms, were also conducted.

Mendelian randomization

Single-nucleotide polymorphisms, which serve as genetic IVs, function similarly to randomized controlled trials (RCTs) in that they are randomly assorted and inherited by offspring, independent of confounding factors. The MR theory is predicated on the basis of 3 key assumptions: (1) Genetic IVs exhibit a strong association with the exposure of interest (P < 5 × 10 − 8); (2) genetic IVs are not correlated with confounding variables affecting the outcome; and (3) genetic IVs do not themselves influence the outcomes under investigation.¹³

All the data used in this study were selected on the basis of previous studies.^14–20 Large-scale genome-wide association analysis (GWAS) data of PCs and MPVs were extracted from the UK Biobank and Investigating the effect of varying the whole blood inter-donation interval (INTERVAL) studies, in which 29.5 million genetic variants were examined for potential associations with 36 red cell, white cell, and platelet properties in a cohort of 173 480 individuals of European ancestry.²¹ The ED GWAS statistics were derived from a recent meta-analysis that included 3 cohorts comprised of 6175 cases out of 223 805 samples. Data from the UK Biobank, the Estonian Genome Centre of the University of Tartu, and the Partners HealthCare Biobank were included.²² The samples utilized in this study were sourced from European populations. The data were employed for MR analysis and are publicly accessible from https://gwas.mrcieu.ac.uk/ (accessed on October 7, 2024).

Genetic instrument selection

The SNPs derived from exposures (PC and MPV) were subjected to filtering on the basis of the criteria of P < 5 × 10 − 8 and linkage disequilibrium r² < 0.01 within a 1 Mb window. The MR-Steiger method was subsequently applied to the selected SNPs to determine the proportion of variance explained by the exposure and outcome variables.²³ The findings of MR-Steiger suggest that the remaining SNPs may have a greater impact on the outcome, specifically ED, than the exposure variables PC and MPV do. Furthermore, to mitigate potential bias arising from weak IVs, the F-statistics of the SNPs were computed via a specific formula.: F-statistics = (Beta/Se)².²⁴ The F-statistic values indicate the magnitude of the IVs, with those below 10 typically considered weak. Ultimately, 220 and 228 SNPs were chosen as IVs for PC and MPV analyses, respectively. These IVs are detailed in Supplementary Tables 1 and 2.

Detection of heterogeneity and pleiotropy

Cochran's Q test was utilized to assess heterogeneity among the genetic instruments. Three methodologies were employed for the identification of pleiotropic outliers.²⁵ The identified pleiotropic outliers were excluded prior to data harmonization and subsequent MR analyses, ensuring adherence to the fundamental assumption of MR in the statistical analysis.

Statistical analysis

Data are presented as the mean ± standard deviation (SD) for parameters exhibiting a normal distribution and as the median with interquartile range for parameters not following a normal distribution. Correlations involving normally distributed data were analysed via Student’s t-test, whereas the Wilcoxon rank-sum test (Mann–Whitney U test) was employed for non-normally distributed data. Univariable and multivariable ordinal logistic regression analyses were conducted for multivariate analysis involving continuous or categorical dependent variables, respectively. Analysis of variance was used to evaluate differences in clinical variables between the 2 populations. The study utilized the IVW method to evaluate the causal effect estimates among PC, MPV, and ED. To increase the robustness of the results, supplementary analyses were performed employing diverse methodologies, including MR-Egger and weighted median methods. The MR-Egger method integrates an intercept term into the Egger regression model. The extent of deviation of the intercept term from 0 can be utilized as an indicator of directional pleiotropy. This method is robust in producing valid causal estimates even when all IVs are invalid. Conversely, the weighted median estimator is capable of identifying causal effects even when up to 50% of the instrumental variables are invalid. The statistical power of different methodologies varies depending on the specific context. The analysis was conducted via R version 4.3.1, with visualizations created via the ggplot2 package. A 2-sided significance threshold of P < .05 was applied to assess statistical significance.

Power analysis

The statistical power of MR was assessed through power calculations via the online tool https://shiny.cnsgenomics.com/mRnd/. A type I error rate of 0.05 yielded a statistical power of 100% for the associations among PC, MPV, and ED. Furthermore, overlap and bias were evaluated via the online software https://sb452.shinyapps.io/overlap/.

Results

Demographic characteristics

A total of 272 participants were included in our study. The patients' demographic information is shown in Table 1. A total of 114 patients reported ED, and 158 reported no ED. The patients in the ED group were older than those in the control group (P = .013). No significant differences in BMI, pulse, blood pressure, low-density lipoprotein (LDL), uric acid (UA), triglycerides (TRY), high-density lipoprotein (HDL), follicle-stimulating hormone (FSH), luteinizing hormone (LH), prolactin (PRL), testosterone (TEST), IPSS, PHQ9, or GAD7 scores were found between the 2 groups. The PC, MPV, Chol, and PDET scores were significantly different between the ED group and the control group (P < .001; P < .001; P = .008; P = .034; P < .001, respectively). Of the factors examined—educational background, residence, city, income, smoking status, and alcohol status—only alcohol status was significantly different between the 2 groups (P < .001).

Logistic regression reveals that PC and MPV are associated with ED

Univariate and multivariate logistic regression analyses were conducted to evaluate various factors. The results of the univariate logistic regression are presented in Table 1. The multivariate ordinal logistic regression analysis indicated that an elevated PC (OR = 1.14, 95% CI: 1.08, 1.22; P = .005) and a reduced MPV (OR = 0.65, 95% CI: 0.48, 0.88; P = .003) were significantly associated with an increased risk of ED, as illustrated in Figure 2. Univariate and multivariate ordinal logistic regression analyses were conducted to examine various factors potentially influencing the risk of ED. Age (P = .013) and PDET score (P < .001) were incorporated as covariates, while the PHQ-9 score, PRL, alcohol status, and cholesterol levels were excluded from the analysis on the basis of clinical experience. The univariate logistic regression analysis indicated a significant association between age (OR = 1.04; 95% CI: 1.01, 1.08; P = .079) and PDET (OR = 0.92; 95% CI: 0.88, 0.95; P = .047) with ED. However, these associations were not observed in the multivariate logistic regression analysis.

Instrumental variable information

After applying stringent criteria for genome-wide significance (P < 5e-8) and standardizing the SNPs, a total of 220 SNPs were identified as IVs for assessing the impact of PC on ED. Following a search in the PhenoScanner database, 9 SNPs associated with risk factors for ED, such as cardiovascular disease, diabetes mellitus, and hyperthyroidism, were excluded, resulting in a final selection of 211 SNPs. In the analysis of MPV and ED, 8 SNPs were excluded, resulting in a total of 220 SNPs for further investigation (Supplementary Table 2).

Causal effects of genetically predicted PC and MPV on ED

The results obtained from different methods are shown in Figure 3a. Genetically predicted PC leads to a 1.09-fold increased risk of ED (95% CI = 1.01, 1.18; P = .016), whereas genetically predicted MPV leads to a 0.93-fold increased risk of ED (95% CI = 0.88, 0.99; P = .014). The scatter plots (Figure 3b and c) reveal that as the SNP effect on PC increases and the SNP effect on MPV decreases, the SNP effect on ED intensifies.

There were no signs of heterogeneity according to the results of the Cochran's Q tests shown in Table 2 (P > .05) or the funnel plot in Figure 4. Additionally, as revealed in Table 2, MR-Egger regression indicated no pleiotropy (P > .05). The results are still robust when any 1 of the SNPs is excluded.

Discussion

Recent research has established a significant correlation between the MPV and ED. A meta-analysis of 14 studies revealed that individuals with ED, particularly those with vasculogenic ED, have higher MPV levels than healthy controls do.⁹ Another study confirmed these findings, showing a standardized mean difference in MPV between ED patients and controls. However, no significant differences were observed in PDW or PC levels between the ED and control groups.¹⁰ Despite these findings, the causal relationships among MPV, PC, and ED remain complex and not fully understood, emphasizing the need for further investigations into the underlying mechanisms and potential confounding factors. In the present study, we performed an observational study to reveal the associations of PC and MPV on ED risk and two-sample MR analysis to investigate their potential causal effects and to determine whether the increase in PC and a decrease in the MPV levels can increase ED risk. The results proved to be reliable through the use of several Mendelian tools to achieve stability in sensitivity analysis. The incidental effects of insomnia,²⁴ snoring,²⁶ and COVID-19²⁷ on ED risk have been well investigated in previous MR studies. Unlike previous studies, this study provides evidence for a potential genetic causality, suggesting that a high PC and a low MPV are associated with a greater risk of developing ED within the context of MR design.

Erectile dysfunction not only has various common risk factors (unhealthy lifestyle, obesity, aging, alcohol consumption, smoking, etc.) but also involves a variety of intrinsic mechanisms, such as endothelial and vascular smooth muscle dysfunction.²⁸ Some studies have shown that increased platelet counts have a greater impact on diseases in which platelets make up a greater proportion of blood clots.²⁹ Numerous clinical studies have shown that PC and MPV play important roles in the pathogenesis of vasogenic ED.^30–34 They may influence ED development by affecting endothelial function and thrombosis. Increased PCs could lead to increased thrombosis, which impacts penile blood flow; whereas, decreased MPV might reflect changes in platelet activation status. These mechanisms may collectively contribute to the onset and progression of ED. Research indicates that platelets play crucial roles in endothelial function and thrombosis, which are key factors in ED development.³⁵ Our genetic analysis further supports this hypothesis, suggesting that platelet parameters could serve as potential biomarkers for ED.

This study has some strengths and limitations. The main advantage of this study is that it is an observational study combined with an MR design, which overcomes endogeneity and bias caused by confounding variables. Considering the challenges of implementing RCTs, this study provides genetic evidence that PCs and MPVs cause ED and pave the way for therapies targeting platelets for ED prevention. However, the sample size of this observational study, which is based on a single center, is relatively small, and the SNP data source was limited to individuals of European ancestry, which limits the applicability of our findings to other ethnicities while avoiding demographic bias. Additionally, the ED summary–level data exclusively present binary variables, leaving the correlation between ED severity on PC and MPV ambiguous. Therefore, future research should prioritize examining the associations between varying degrees of ED severity and platelet parameters. Finally, the available PC and MPV data are not sex specific. Conducting sex-specific analysis could provide more detailed information in the future.

Conclusion

In summary, this research utilized MR to examine the correlation between genetically determined PC, MPV, and ED. Our findings revealed that genetically determined elevations in PC and reductions in MPV are associated with a heightened risk of ED. Consequently, comprehending the associated risk factors is a critical concern in both clinical practice and public health. These results have significant implications for the prevention, detection, and management of ED.

Acknowledgments

We acknowledge the HOME for Researchers platform for its provision of a convenient drawing instrument.

Funding

This work was supported by the Hunan Provincial Natural Science Foundation (S2024JJQYLH09).

Conflicts of interest

None declared.

References